Vacuum Web Coating

Problem with Rootes Blower

What could be the cause of a roots blower pressure to plunge intermittently?

Answer

There are some questions and checks that you can do.

Is the same pressure change seen by more than one pressure gauge?

If the pressure variation is seen only on one gauge then does the noise the Rootes pump makes change with the change in pressure indicating that the extra gas load is real or if you measure the current to the pump does this change with the change in pressure also indicating a real change in gas load.  If the sound or current does not change then it might mean that the pressure gauge is faulty rather than the pump.

Assuming it is not a faulty gauge the ask what has changed between the pump not having a problem and the problem appearing?

This could range from nothing at all to things such as the system being overhauled, or maintained where water or air joints have been undone and re-made, or the system has been greased/oiled as part of a routine maintenance.  If something has been done to the system then this could be the source of the problem. Whatever has been done may have disturbed something such as a seal is not longer sitting properly and so is leaking, etc

Check the simple things such as the oil level in the pump, has the correct oil been used, and check the pressure gauges are reading correctly. 

Having done all this then there are two basic possible problems.

The first is that there is a problem with the vacuum. This could be a water leak where the water is leaking at a low rate into the vacuum such that as it evaporates into the vacuum the heat required to evaporate is taken from the water and pipe and this is cooled to the point where the water freezes. The ice then has a lower vapour pressure and so the vacuum pressure falls. As the evaporation rate has fallen this takes less heat and so the thermal mass begins to heat the ice and it returns to water and so the higher rate evaporation returns and the pressure rises. The cycle then repeats and the water freezes and the pressure falls again.  This generally only occurs for relatively small leaks and the two pressures are characteristic of the vapour pressure of water and ice.

There could also be a small air leak and a pipe that is deforming so that sometimes the leak is closed and other times open.

The second possible type of problem is that there is a problem with the Rootes pump.  This could relate to gas loading. If the gas load is too high the rotors could be bending and touching the housing causing an increase in friction and so slowing down the rotation speed. Other problems could be a problem with the seals that is causing an intermittent leak.

I hope this gives you a few things to think about and try.

Pumping problem

I have 2 metallisers in my company.
We have bought the 2nd one recently from a company in USA. It was never used before, so it was practically brand new. We have been operating the first system for a while now. I am facing a typical problem. I am loosing a lot of diffusion pump oil from the second machine. On an average this can be as much as 1 gallon a week. We utilize one charge in a year in the first system. When I asked the company about it they say there are 2 reasons.         1. Poor foreline vacuum or        2. Leakage.

I have checked both and am not able to figure out where the problem is, because the readings are as good as normal. Could you please help me out with this?

Answer.

The poor foreline pressure would lead to backstreaming, particularly at low pressures.

The backing pump is not matched to the diffusion pump requirements and so the diffusion pump pressure rises and this causes some of the oil to backscatter and enter the main vacuum chamber.

What you should be looking for is evidence of oil in the main chamber.

Generally most vacuum systems will have a small amount of oil that can leave a thin smear of oil on the surface but if you are losing 1 gallon per week then I would expect that there would be much more pronounced evidence of oil further into the chamber and possibly even signs of oil drips on the bottom edges of some of the metal surface nearby.

If you have any baffle across the diffusion pump orifice I would expect it to be well covered in oil. If this baffle is also your cryo coil then I would inspect the water that you collect when the system is brought back to atmosphere and see if it has oil included in the water.

The problem of leaking would work the other way round. The gas load is so great on a permanent basis that the diffusion pump oil is being swept out and through the backing pumps.  This might be noticed if the backing pump oil levels have increased and possibly they are throwing a lot more oil out in the backing pump exhaust.  If you have a backing line with a low spot in the tubing where oil could collect, it may be worth checking to see if oil is accumulating there. However several gallons I would have expected would have become evident by now and so I would put this as a lower priority than other checks.

Both of these problems would suggest that the system base pressure is not as good as it could be. A poor foreline pressure would mean the diffusion pump would never reach its ultimate base pressure. A leak also would not allow it to reach its ultimate pressure. A quick check for a leak is to pump the system down and then seal it off and measure the pressure rise with time. If the pressure rise continues with a reasonably constant rate of rise in pressure with time then the system has a leak, with the higher the slope the greater the size of leak. If the pressure rise is initially quick but then becomes slower and levels off then the system does not have a major leak and you need to be looking to a problem with the backing pressure.

If you purchased the system that was aimed for a different location it is worth a check that it is not a different voltage problem that is making the backing pumps not perform as well as expected.  Sometimes the change in voltage can mean they are operating below full capacity and many system designers minimise the cost of the machine by using the minimum pumping they can get away with and so a small drop in performance can radically affect the pumping performance.

Other checks on the backing pumps are the standard ones, check the oil level is correct and that the pumps have not collected too much water and need gas ballasting. If the oil level and condition is OK then the same pressure check can be done with the backing line. Pump out the backing lines and then seal them off and check the pressure rise. It will be the same as for the main chamber, if the pressure continues to rise you have a leak but if there is an initial pressure rise and then the rise slows and stops then there is not leak.

Also check the backing pumps to see if the oil is one of the recommended ones for the pump. Sometimes oils can be sold cheaply because they do not have the same operating performance. 

I think that this should give you a few things to try.

Will this improve the pumpdown speed?

Question

We have a metalliser that has two deposition zones. The volume is very large (15.53m 3). We are not use currently the second deposition source so we think to put up an empty airtight box in the empty space. Is there any benefit for the vacuum or is there any disadvantage in the sense of degassing of the metal surface?

Answer.

Essentially what you propose is possible and by reducing the volume you will reduce the pumpdown time.

There are a couple of points to bear in mind when making this type of change.  The box MUST be leak tight otherwise you will have a virtual leak permanently within the vacuum system and this will both reduce the pumpdown speed and the base pressure.  Depending where the volume is positioned will depend on if it collects any stray deposition. If it does it needs to be cleaned like any other surface to keep the surface area down to a minimum, otherwise you will be absorbing more water than before and this too could reduce the pumpdown speed, reducing the benefit. The increase in surface area of just the box surface is small compared to the surface area that is produced by any stray coating and so generally would not significantly change the amount of outgassing that would occur during the pumpdown. It is only if it collects stray deposition and it is difficult to clean and so gets left with the coated material on that it would affect the pumping.  Finally check the position with respect to any conductance gaps and the positioning of the pumps. You do not want the box to block any pumping orifice. If the box reduces the size of the pump throat then this would act as a constriction reducing the pumping speed of the orifice and slowing down the pumping of the vessel.

Assuming the box is leaktight, kept clean and suitably positioned then you can expect a pumping speed improvement.

Rootes pumps and Booster pumps

What is the difference between the roots pump and the booster pump?

ANSWER

Booster pumps are usually a similar type of pump to a diffusion pump in that some oil is vaporised and directed down a reducing volume to help knock the gas down the tapered tube to compress the gas and so increasing the pressure to make it easier to pump away by the backing pumps.

The rootes blower to some extent does the same job but it is usually classed as a backing pump and is used to improve the crossover as the backing pumps (often rotary vane pumps) run out of pumping speed and before the diffusion pump becomes effective. Often the rootes pumps are place in-line with the roughing line. The rootes pump is often left to freewheel without being driven when the chamber is initially roughed down. Once the pressure falls sufficiently the rootes pump starts to be driven where the rotating lobes can cause a reducing volume forcing the pressure up and making the rotary pumps more effective.

Hence using either can make the operation of the total system more efficient.  Modern vacuum systems appear to favour the rootes plus backing pump combination over the combined diffusion pump plus booster pump combination.  I must confess I have never checked on why, it could be that the cost benefits or speed is better with the rootes pump than with the booster. 

Question re. condensing gases onto cryocold surfaces

Question.

At the moment we are using a PolyCold and cryogenic panel to remove retained solvents on 12μ polyester in our metallizer.

This works ok on a web of 700mm, but will not work on a web of 1400. Would it be better to use liquid nitrogen as a cooling medium instead. (not sure as to what temp. solvent condenses at).

*************************************************************

Answer.

When you talk about polyester film and using a cryogenic panel to remove retained solvents I am curious about your substrate material.  Usually polyester does not have much if anything by way of retained solvents. However it does contain moisture and depending upon the manufacturing process and then the subsequent winding and storage conditions this can be as much as 2% by weight of moisture.  Most Polycold cryopanels are used specifically to pump water vapour and nothing else.

The way any cryopanel works is to lower the temperature of the panel to the point where any atoms hitting the surface have a high sticking coefficient. The temperature governs what atoms will stick and which will not. The Polycold type cryopanels have the temperature reduced enough such that they have a very high sticking coefficient for water molecules but little else.  Gases such as oxygen or hydrogen are generally only captured if the water vapour traps the gas as it condenses and solidifies into ice.

The performance of the cryopanel depends upon the capacity of the system. As the gases condense and solidify an ice layer builds up. Ice is not as good a thermal conductor as the copper tube or panel surface. The ice may not form as a dense layer as the density depends on the rate of arrival of the gas and if the rate is very high the layer is likely to be porous. A porous ice layer is an even worse thermal conductor. The arriving gas is at a higher temperature than the cryopanel and so the cryo panel will warm up. The amount it warms is dependent upon the gas load and the capacity of the cryopump to keep up with the gas load.  The capacity of the cryopump has to be sized to pump the last amount of water vapour from a roll every bit as well as the first few meters otherwise there will be a progressive pressure rise through the process as the pumping performance declines and the ice layer builds up and the porosity gets progressively worse.

Thus on your system if the cryopanel has been sized to cope with the gas load from a 700mm wide roll of polyester and it then has in effect double the load applied then it is likely that the cryopanel temperature will rise to a higher level during the process and the ice layer formed is also likely to be less dense because of more gas arriving which means the pumping performance of the cryopanel is likely to decline.  If the cryopumping system has the capacity it may be that simply doubling the surface area of the cryopanel will cure the problem.  If the cryopanel surface is already at a maximum it may be that you need a second cryopump.

Worried by backstreaming?

Backstreaming is a common concern and yet many operators still insist that vacuum systems need to be pumped to a very low base pressure before starting the deposition process.  These two facts seem to be contrary, as we all know that backstreaming becomes increases with reducing pressure & time spent at the low pressure.

Therefore it seems only sensible not to pump down to a low base pressure.

However if I suggest this method the immediate response is what about the water vapour absorbed on the walls?  The belief is that pumping down to a low base pressure helps remove the water from the walls. 

            This is not true; the ability to pump to a low pressure is an indication that the water has been removed. The lower pressure does not actually remove the water.

There are several ways of speeding up the removal of the water. One is to use heat and this is commonly used on ultra-high vacuum systems with trace heating elements wrapped around all the external pipe-work. The second method is to apply energy by some other means such as ultra-violet lamps that excite the water & assist in breaking the bond to the surfaces.

            The final method, that has been little used, but works is to only pump to a set pressure which can either be where the roughing pumps are still working or to a point just below the stalling point of the diffusion pump.  At the elevated pressure there is increased bombardment of all surfaces because of the higher gas pressure.  The greater number of gas molecules means that there are more molecules bouncing off the walls & the kinetic energy is sufficient to break the bonds holding the water to the surfaces & hence it speeds up the release of the water.

            Thus feeding the system with a small leak of dry nitrogen for a set time is better at reducing the water than trying to pump to a lower pressure. If there is dry nitrogen in the system it may be possible to strike a plasma which will further increase the energy & so will further increase the water evolution rate.

            The nitrogen, or other inert gas, needs to be dry otherwise the process is counterproductive.

            By this method it is possible to reduce the pumpdown time slightly as well as never having to pump to the ultimate base pressure and risk backstreaming.  In turn this will them improve the system productivity & the film quality.

  If in doubt try it out   -    & let me know how you get on.

Common Pumping Problems.

Water has to be the biggest cause of pumping problems.  The source of the water can vary from, a wetter than normal substrate, more ambient moisture absorbed when the machine is at atmospheric pressure or even a water leak directly into the system.

Other problems can relate to a lack of routine maintenance or too infrequent maintenance.  This can be everything from not cleaning the deposition shields frequently enough to not checking the oil levels in the pumps or ballasting the roughing pumps to remove any excess water vapour.

Below I will highlight each of these problems and some of the methods for identifying the possible source of the problem.